Project Details
Description
Following International Accounting Standards, this project is realized by the Institute for Information Business and the Research Institute for Computational Methods in cooperation with its corporate partner PARAMETRICA, a leading financial and actuarial advisory company in Italy. This project estimates the future pension liability that a company has to build for its employees.
To estimate this future pension liability, the application uses the actuarial methodology of Projected Unit Credit Method (PUCM) within a stochastic simulation framework.
The team HPCFaI (High Performance Computing Finance and Insurance implemented an application based in JAVA and C that after specification of economic and actuarial hypotheses (interest rates, the inflation rate, predicted growth rates for the salary of the employees, hypotheses about death-probabilities and probabilities of invalidity and resignation), estimates the future pension liability of a company with the Monte Carlo Simulation approach. The result of that calculation is used directly in the balance sheet.
The core technology is JAVA, but computing intensive parts of the applications are implemented in C. For the communication between JAVA and C the team decided to use the Java Native Interface (JNI).
The universitys infrastructure, especially the Cluster@WU offers the team the ability to gain enormous reductions in computational time. In the future it would be possible to run a huge number of simulations and estimate lots of scenarios within a reasonable computational time.
To estimate this future pension liability, the application uses the actuarial methodology of Projected Unit Credit Method (PUCM) within a stochastic simulation framework.
The team HPCFaI (High Performance Computing Finance and Insurance implemented an application based in JAVA and C that after specification of economic and actuarial hypotheses (interest rates, the inflation rate, predicted growth rates for the salary of the employees, hypotheses about death-probabilities and probabilities of invalidity and resignation), estimates the future pension liability of a company with the Monte Carlo Simulation approach. The result of that calculation is used directly in the balance sheet.
The core technology is JAVA, but computing intensive parts of the applications are implemented in C. For the communication between JAVA and C the team decided to use the Java Native Interface (JNI).
The universitys infrastructure, especially the Cluster@WU offers the team the ability to gain enormous reductions in computational time. In the future it would be possible to run a huge number of simulations and estimate lots of scenarios within a reasonable computational time.
| Status | Active |
|---|---|
| Effective start/end date | 1/07/07 → … |